Identification of a typical CD player arm using a two-layer perceptron neural network model

This paper investigates the identification of a typical CD player arm using a two-layer multi-layer perceptron neural network. It is shown that the neural network based state-space innovations form model clearly outperforms the equivalent linear model in simulation and thus, it is possible to obtain good results for this system with neural network based state-space innovations form model. It is by no means claimed that the "optimal" solution is found, however, it is shown that the proposed neural network based model has provided a simple means to solve the given nonlinear multi-input-multi-output (MIMO) system. In addition, as correlation functions of the prediction errors tend to remain confined more closely to the confidence regions, it is likely that most of the information has been extracted from the training set and that the neural network based model tries to approximate the system fairly well.